Journal Pre-proof

Simulation, statistical modeling, and optimization of CO2 laser cutting

process of polycarbonate sheets

Mahmoud Moradi, Mojtaba Karami Moghadam, Mahmoud

Shamsborhan, Zeinab Malekshahi Beiranvand, Alireza Rasouli,
Mahdi Vahdati, Arash Bakhtiari, Mahdi Bodaghi

PII: S0030-4026(20)30768-3
DOI: https://doi.org/10.1016/j.ijleo.2020.164932
Reference: IJLEO 164932

To appear in: Optik

Received Date: 6 March 2020

Revised Date: 9 May 2020
Accepted Date: 10 May 2020

Please cite this article as: { doi: https://doi.org/

This is a PDF file of an article that has undergone enhancements after acceptance, such as
the addition of a cover page and metadata, and formatting for readability, but it is not yet the
definitive version of record. This version will undergo additional copyediting, typesetting and
review before it is published in its final form, but we are providing this version to give early
visibility of the article. Please note that, during the production process, errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal
pertain.

© 2020 Published by Elsevier.

Simulation, Statistical Modeling, and Optimization of CO2 Laser
Cutting Process of Polycarbonate Sheets

Mahmoud Moradi 1, 2, *, Mojtaba Karami Moghadam 1, 2, Mahmoud Shamsborhan 3, 4,

Zeinab Malekshahi Beiranvand 5, Alireza Rasouli 1, 2, Mahdi Vahdati 6, Arash Bakhtiari
1, 2
, and Mahdi Bodaghi 7, *

1 Department of Mechanical Engineering, Faculty of Engineering, Malayer University, Malayer, Iran

2 Laser Materials Processing Research Centre, Malayer University, Malayer, Iran
3 Department of Engineering, Mahabad Branch, Islamic Azad University, Mahabad, Iran

4 Department of Mechanical Engineering, University of Zakho, Kurdistan Region, Iraq

5
Department of Materials science and Engineering, Tarbiat Modares University, Tehran, Iran

of
6 Faculty of Mechanical and Mechatronics Engineering, Shahrood University of Technology, Shahrood, Iran

7 Department of Engineering, School of Science and Technology, Nottingham Trent University, Nottingham, United

Kingdom

ro
*Corresponding authors: mahdi.bodaghi@ntu.ac.uk (M. B.); moradi@malayeru.ac.ir (M. M.)

Highlights:
-p
re
 Simulation of CO2 laser cutting of polycarbonate injected sheets by the DOE
approach

lP

Developing governing equations on thermal fields and boundary conditions

 Modeling of the heat source by selecting of an appropriate Gaussian distribution
 Laser power, focal plane position and cutting speed considered for the DOE analysis
na

 Optimizing the laser cutting simulation process

Abstract
ur

Laser cutting well-known as a manufacturing process is a rapid, repeatable, and reliable

method that is frequently used for cutting various materials such as thermoplastics. Due to
Jo

their physical and chemical properties such as fatigue resistance, high toughness, and re-
melting properties, thermoplastics such as polycarbonate are widely used in automotive parts,
electronics, etc. In this study, a numerical simulation of the laser cutting process by a finite
element method is developed. The sample simulated in this research is a 3.2 mm thick
Polycarbonate sheet that is subjected to the laser cutting process by a low power continuous
CO2 laser. The effects of the laser cutting process parameters such as laser power, cutting
speed, and laser focal plane position on the top and bottom kerf width, top heat-affected zone,
the ratio of upper kerf width to lower kerf width and taper kerf are investigated by statistical
techniques of variance analysis. Choosing an appropriate Gaussian distribution is studied as

1
well. The results show that the laser scanning speed has a significant effect on the top kerf
width. By choosing a cutting speed of 20 mm/s and a focal length of -3, the taper kerf is
minimized. By increasing the laser cutting speed from 4 to 20 mm/s and decreasing the laser
power from 50 to 20 Watts, the heat-affected zone decreases. The developed analysis can
predict the depth of kerf in a continuous mode for different values of laser power, speed, and
laser focal plane.

Keywords: polycarbonate; laser cutting; optimization; finite element method; statistical

modeling; experimental validation

of
ro
-p
re
lP
na
ur
Jo

2
Introduction
Laser cutting is a two-dimensional machining process that involves the removal of materials
by high-intensity beam focusing on the workpiece. The laser beam melts or evaporates the
workpiece in the thickness or depth direction of materials and creates a kerf wall. Then the
molten material is blown away using an assistant gas. In the laser machining, unlike
traditional methods, material-removing is almost conducted without applying any force. This
feature makes the laser machining suitable for delicate and brittle parts on which too much
force should not be applied. Ultra-thin kerf width, which saves a significant amount of
materials, is another feature of the laser cutting. In the laser machining, the final surface
quality can be achieved in one step, and no further finishing operations are needed to smooth
and clean the surface. Minor noise pollution, shallow heat-affected zone, very high cutting
speed, and ease of automation are the other features of the laser cutting. Lasers are capable of
cutting a wide range of brittle or ductile, conductive or nonconductive, hard or soft materials.
Table 1 compares the laser cutting process with the other cutting processes from different

of
aspects.

Table 1- Comparison of different cutting process types

ro
Quality Laser Wire cut EDM Ultrasonic Abrasive fluid jet Plasma
Speed     
Kerf quality     
Kerf width
Wear of tool






-p







Metal and non-metal
re
Heat-affected zone     
lP

Polycarbonate (PC) is a transparent polymeric material from the thermoplastic group that is
usually made in the form of plates with a flat or dentine surface. This material has excellent
optical properties and high impact resistance [2]. The impact resistance of the polycarbonate
sheet is 250 times higher than a glass with the same thickness. Various sheets of
na

polycarbonate show scratch resistant like glass, under very tough test conditions. Among the
advantages of polycarbonate coating in comparison with glass are its lower cost and lighter
weight. Furthermore, its high resistance in contrast to plastic has led to its increasing demand
in the industry. Chemical resistance to many organic acids and mineral oils, excellent
ur

flexibility, and low thermal conductivity (suitable thermal insulation) have led to the
increasing expansion of polycarbonate. Polycarbonate is highly resistant to Ultraviolet (UV)
rays due to its anti-UV coating. Its color also does not change in the sun’s radiation,
Jo

consequently retains its transparency for a long time. These properties have led to the use of
polycarbonate in the automotive, health, and medical, electrical, and computer industries. So
far, various studies have been done on the laser cutting process. These studies involve
experimental activities on metallic and non-metallic materials [3, 4, 5, 6, 7]. In many of these
studies, process modeling has been done using statistical methods and neural networks. Due
to the problems and costs of experiments, in many other studies, the process modeling has
been done by numerical and analytical techniques.

3
By studying previous studies, a set of parameters and quantities evaluated in different studies
are shown in Figure 1. Then, the latest research on laser cutting of polymer materials is
described.

of
Figure 1 – Aspects and quantities investigated in the previous studies on the laser cutting process [8].

ro
In 2019, Aref Varsey et al. [9] studied the taper kerf resulting from laser cutting on
Polymethyl methacrylate (PMMA), using statistical methods. In this study, the effect of laser
power parameters, the scanning speed, and the number of sweep steps was investigated using
-p
analysis of variance. Experimental studies using statistical methods on the cutting processes
were carried out by Haddadi et al. [10]. Chen et al. [11] used an experimental design
technique (Taguchi method) to enhance the surface quality of microchannels obtained by
re
laser cutting on PMMA materials. In this study, power parameters, the scanning speed, and
the time of process were optimized to achieve less roughness. They were able to make an
average roughness of 110 nm. Saleh Jassim et al. [12] conducted studies on the cutting and
lP

drilling of the high-density polyethylene pipes by CO2 laser. In this research, the effect of
injected gas during the process, laser power, and the thickness of the workpiece on the kerf
width and heat-affected zone were evaluated. Gruska et al. [13] investigated the effect of
laser power and laser speed on the surface quality of polycarbonate gears by the Taguchi
na

method. Hoffey et al. [8] summarized the results of research on laser cutting on fiber-
reinforced polymer composites. The results show that most studies are focused on the heat-
affected zone, width, and depth of kerf. This study demonstrated that short pulses, high
ur

speed, low laser intensity, and high pressure of injected gas should be used to achieve the
most moderate heat-affected zone. In 2018, Aref Varsi et al. [14] designed algorithms to
predict the number of laser scanning steps in laser cutting on PMMA materials to achieve a
Jo

specific kerf depth. There was a good agreement between the result of the algorithms and the
experimental results.

In the engineering fields, different methods are used for tackling problems with high
accuracy, and one of the most practical approaches for engineers is finite element method
(FEM). It is employed in different areas such as heat transfer, fluid flow, structural analysis,
electromagnetic potential, and mass transport. There is a large number of experimental
studies on laser cutting, but analytical studies on the nature of this process and its parameters
are very limited. Recently, the use of the finite element method for simulating laser cutting

4
has been expanded. In many studies in this field, the temperature distribution resulting from
the laser beam on the surface of the workpiece (metal and non-metal) has been analyzed [15-
19]. In some studies, the kerf depth, taper kerf, and surface quality of kerf have been
investigated, too [20-23]. In 2019, Benton et al. [24] evaluated the effect of laser
micromachining parameters and thermal properties of the workpiece on the kerf depth
resulting from this process, for the fabrication of microchannels on polymeric PMMA
materials by finite element methods (COMSOL software package). In 2019, Jiang Nam Li et
al. [25] used a numerical solution method (COMSOL software package) to determine the
effect of the laser speed and laser power on the laser cutting process of PMMA materials. In
this modeling, the Arbitrary-Eulerian-Lagrangian method technique was used. Moradi and
Golchin [26] chose a suitable Gaussian distribution for modeling in the laser drilling
simulation. New combined techniques with Laser machining have been developed in recent
years to reduce the thermal damage and prevent the formation of the resolidified layer. The

of
laser micromachining method on the ice layer is one of the new techniques. In 2019,
Macklow developed a thermal model to examine this technique more accurately [27]. Also,
in 2019, Choi et al. [28] presented a three-dimensional thermo-elastic model to understand

ro
the method proposed to solve the problem of laser cutting of brittle materials. One of the gaps
in previous studies that have been examined in the present study is choosing a suitable
Gaussian distribution to simulate laser cutting. Furthermore, numerous brilliant studies were
-p
conducted by FEMs in many fields [29-34]. Reviewing previous numerical studies shows that
statistical methods have not been used in these studies to investigate the effect of parameters
and extract the regression model. Table 2 illutreates some studies on cutting processes
re
investigated by various methods.

In the present study, the effects of process parameters including cutting speed, laser power,
lP

focal length on the kerf width, the top and the bottom kerf width, taper kerf, and heat-affected
zone are investigated by analysis of variance. The optimal condition for each output is also
determined. Due to the polycarbonate properties, the laser cutting process results in color
opacity (yellowness of plane) as well as burnt edges. Appropriate choosing of laser cutting
na

process parameters increases the kerf quality as well as eliminating burnt edges in this
process. The widespread use of polycarbonate in the industry highlights the importance of
this study.
ur
Jo

5
Table 2- Different cutting processes applied in various studies

Authors Materials Method Goal(s) Year Ref.

Polymethyl Statistical Parameters
Varsi et al. 2019 9
methacrylate methods optimization
Improving the
Response
Haddadi et al. Polystyrene Laser cutting 2019 10
surface method
condition
Orthogonal
Polymethyl Laser cutting
Chen et al. experimental 2017 11
methacrylate optimization
method
Thermal
conductivity
Saleh Jassim et Experimental
Poly ethylene and polymer 2019 12
al. Study
absorption of
the (CO2) laser

of
Model for
Polymer Matrix Finite Element
Sihn et al. Temperature 2019 15
composites Method
Evolution

ro
Analysis of
Inconel 803 Finite Element
Yilbas et al. Thermal Stress 2019 17
Alloy Method
Field
Simulation of
Amara et al. Stainless steel
Volume-of-
fluid method
-p
the workpiece
kerf formation
with CO2 laser
2016 18
re
Controlling the
Ti6Al4V Finite Element
Ayed et al. reduction of the 2014 20
titanium alloy Method
cutting force
lP

Simulation of
Finite Element
Meško et al. Low alloy steel laser cutting 2018 21
Method
process
Simulation of
Volume of fluid Transient 3D
na

Kheloufi et al. Mild steel 2015 22

method Temperature in
Laser Cutting
Arbitrary
Effects of laser
Polymethyl Lagrangian
Li et al. machining 2019 25
ur

methacrylate Eulerian
parameters
method
The mechanism
Finite Element
Choi et al. silicon wafer of crack 2019 28
Jo

Method
propagation
Defect
characteristics
Carbon fiber
Finite Element and strain
Li et al. reinforced 2020 35
Method distribution
plastic
following fiber
laser cutting
Simulation of
Finite Element
Kaselouris et al. AISI H-13 Steel laser cutting 2020 36
Method
process

6
2- Finite element simulation
Due to many parameters affecting the laser cutting process, changes in material properties
during machining resulting from thermal variations, and other complexities in the process,
achieving accurate estimation of the amount of material removal is relatively complex and
challenging and analytical methods can only be done with a lot of simplifying assumptions.
Therefore, a finite element simulation of the laser cutting process can be considered as an
appropriate method to predict the thermomechanical behavior of this process.
The aim of thermal modeling is the calculation of thermal distribution in the kerf of the
sample. Therefore, all the thermal effects of the workpiece, the laser, and gas jet are involved
in the results. Also, the thermal parameters of the workpiece material are determinant in the
thermal field results. To create an accurate thermal model for the laser cutting process, the
following considerations are necessary:

of
- Heat transfer from the surface of the melt pool and the workpiece to the surrounding
environment is carried out by two mechanisms of convection and radiation.
- Large thermal gradients occur in the melt pool.

ro
- Thermal and mechanical properties of steel, including heat conductivity, specific heat,
density, Young’s modulus, yield stress, and plastic strain, have been used as a function of
temperature.
-p
- To consider the effect of heat loss resulting from gas jet, the convection coefficient of 1000
W/m2k is applied for kerf zone.
- Specific latent heat is considered as an incremental amount of specific heat in the phase
re
change region.
- The laser beam is modeled as a heat source with Gaussian-distributed heat flux.
lP

2-1- Governing equations on thermal fields and boundary conditions

The basic equation of thermal conduction denotes the thermal distribution in the workpiece
based on the law of energy conservation, the capacity of the workpiece to store this energy,
na

and the heat transfer rate across the boundaries according to the Fourier law. The three-
dimensional heat transfer in the Cartesian coordinates is expressed as equation 1 [37-38]:
ur

(1)

where kx(T), ky(T) and kz(T) represent non-isotropic thermal conductivity coeficients, ρ(T) is
Jo

heat-dependent density, C(T) is heat-dependent heat capacity, and t is time. For an isotropic
material, the thermal conductivity is equal, along with all three coordinate directions.
Equation 1 is rewritten in a matrix form as equation 2:

(2)

7
As stated in the assumptions, heat transfer at the surface of the workpiece is considered as
thermal convection and radiation. By solving the equations governing the transient heat
transfer in the workpiece, with determined boundary conditions and certain initial conditions,
the temperature field in the model is obtained. In the laser cutting, because of the transient
process and mass removal phenomena, as well as having a wide temperature range coupled
with the phase changes and the dependence of material properties on temperature, these
equations take on a very complicated nonlinear form that solving them through mathematical
relationships is not easy. Thus, for modeling of the laser cutting, a finite element method will
be developed.

2-2- Simulation method

First, the thermal model is presented to calculate the heat-affected zone. Then the

of
measurement of the top and bottom kerf of the workpiece resulting from laser radiation, will
be described. After applying heat flux to the surface of the workpiece, elements whose
temperature is higher than the melting temperature of the material are removed from the

ro
workpiece surface.
To simulate this process, the element birth and dead technique is used. That means elements
-p
whose temperatures are higher than the melting temperature, are death. Figure 2 shows the
three-dimensional cross-section of the geometrical model with dimensions of 80*175*3.2
mm. It is obvious that the side of the workpiece, which is influenced via laser treatment, has a
re
fine mesh. Due to the high-temperature gradients near the kerf zone, the number of elements
in this region is higher than the number of elements in the farther points from the kerf zone.
Connecting the fine elements to the large elements is done through pyramid-shaped
lP

intermediate elements.
na
ur
Jo

Figure 2- A view of mesh detail of the model geometry

The material used in this study is a colorless U1100 granola polycarbonate sheet with
temperature-dependent thermal properties. Therefore, in the simulation of heat transfer
processes, it is necessary to apply these properties as a function of temperature. The
temperature-dependent thermal properties of polycarbonate are presented in Table 3.

8
Table 3- Physical and thermal properties of polycarbonate U1100 at different temperatures [7].
Coefficient of thermal
Temperature Density Specific heat conductivity
)°C( (g/cm3) (j/g.k) (w/m.k)
25 1.193 1.32 0.199
43 1.191 1.33 0.203
69 1.188 1.35 0.206
93 1.185 1.37 0.211
118 1.181 1.39 0.215
142 1.170 1.47 0.212
167 1.512 1.49 0.211
192 1.138 1.50 0.213
221 1.111 1.53 0.222
246 1.097 1.53 0.226
271 1.070 1.550 0.232

of
298 1.057 1.557 0.237

2-2-1- modeling of the heat source (selecting of an appropriate Gaussian distribution)

ro
In the laser cutting simulation, the Gaussian distribution is often used as an optimal
distribution for modeling the laser heat flux. The Gaussian distribution has a unique
geometric shape, but the Gaussian distribution presented for simulating laser cutting may be
-p
different in some parameters. In this study, three different Gaussian distributions are used for
simulation and their results are compared with experimental results [2].
To obtain the appropriate Gaussian distribution, three cases of experimental results are
re
considered as a baseline, and three different Fortran codes are written for the three different
Gaussian distributions. Simulations are then performed under conditions of these experiments
for each of the Gaussian distributions. Due to the difference between the results of the first
lP

Gaussian distribution and the experimental results, the suitability of the Gaussian Distribution
No. 1 is confirmed. Since the ABAQUS software is not capable of moving the heat source
onto the model as well as the heat flux distribution in the Gaussian form, an ABAQUS Dflux
subroutine in the Fortran programming language is developed in this study. Applying
na

Gaussian heat flux results in equation 3 and 4 as:

(3)
ur

√ (4)
Jo

where p is the absorbed laser energy, r0 is the radius of the laser beam, r is the radial distance
from the laser beam axis, H is the thickness of the workpiece, and y is the current coordinate
in-depth direction. The radial distance from the laser beam axis is calculated from equation 4,
where v is the laser velocity, zin is the initial coordinate of the laser in cutting direction, x and
y are the current coordinates at any time, during the laser movement [39]. Figure 3 also shows
an overview of the Gaussian distribution used in this study. The peak of Gaussian distribution
has a significant effect on the laser cutting process because it has a high temperature in this
area. By this property, the workpiece is cut very smoothly.

9
Figure 3- Gaussian distribution of the laser beam

The simulated laser radiation has a maximum power of 50 Watts and a wavelength of 1064
µm. In the thermal model, DCC3D8 elements are used. This element, is a rectangular cubic
8-noded element with a linear connection whose only degree of freedom is temperature, and

of
it can be used in both stable and transient thermal analyses. Due to the symmetry of the
model along the kerf line, the heat transfer is also symmetrical to this line, and there are no
boundary conditions for it. In other words, it is considered isolated. For other outer

ro
boundaries of the sheet, convection and radiation boundary conditions are used. The radiation
coefficient and the convection coefficient are considered equal to 0.6 and 30 W/mK,
respectively. For the floor of the workpiece that is attached to the ground, the convection
-p
coefficient is 100 W/mK. The ambient temperature is also considered to be 25 °C. Element
birth and death technique is used to simulate the material removal during the cutting process.
This technique is based on deactivating elements of the melt pool during cutting. Thus, by
re
starting machining and applying the heat flux, the elements whose temperature exceeds the
melting temperature of the material are deactivated, and do not participate in the calculation.
Top kerf width, bottom kerf width, alpha angle, and heat-affected zone are investigated as
lP

simulation outputs. How to measure the above parameters is shown in Figure 4. According to
Figure 4, the alpha angle, which is shown as α in this figure, is calculated from equation 5
defined as [2]:
na

(5)
ur

In this equation, Bottom kerf and Top kerf are changeable. Figure 5a shows an overview of
the bottom of the model after the thermal simulation. The heat-affected zone of the workpiece
is calculated from the thermal analysis. Based on Figure 5a and thermal analysis, laser heat is
Jo

more effective on top of the workpiece, and this phenomenon happens because of the shape
of Gaussian distribution. As mentioned before, the mesh size in this zone is very smaller than
the far part of laser treatment, and it helps to increase the accuracy of the calculation in this
area. Figure 5b also illustrates an overview of the model after the death of the elements where
the top and the bottom kerf width are calculated from this analysis.

10
Figure 4- Schematic of the geometrical profile in the cross-section of kerf [2].

of
ro
(b) (a)
Figure 5- a) The view of the bottom of the model after thermal simulation, b) the view of the bottom of the
model after the death of the elements.
-p
A comparison of the experimental results [2] with three different Gaussian distributions is
re
presented in Table 4. Due to error values obtained from comparing calculations with
experimental results for all three Gaussian distributions, the Gaussian Distribution No. 1 is
used to perform other simulations.
lP

Table 4- A comparison between experimental results and different Gaussian distributions in the
simulation
Gaussian Gaussian Gaussian
No.

Laser Powe
W

distribution #1 distribution #2 distribution #3

na (experimental)

(experimental)
Cutting speed

Bottom kerf

Bottom kerf

Bottom kerf
Bottom kerf

Error %

Error %

Error %

Error %

Error %

Error %
Top kerf

Top kerf

Top kerf
Top Kerf
mm/s

FPP
mm
ur

1 6 35 -3 265 163 13 8 22 19 19 19
2 10 30 -3 312 163 12 11 25 21 18 17
3 10 20 0 306 287 10 7 27 24 16 15
Jo

2-3- Validation of the model

After selecting the appropriate Gaussian distribution and creating the model to simulate the
laser cutting process of polycarbonate, the model is validated using experimental results
reported in Ref. [39]. A comparison between simulation and experimental results is presented
in Table 5.

11
Table 5- The comparison between simulation and experimental results
Inputs Top kerf Bottom kerf
No.

Experimental

Experimental
Percent error

Percent error
Simulation

Simulation
FPP (mm)
(mm/s)
Power

Speed

(μm)

(μm)

(μm)

(μm)
)%(

)%(
(P)

1 35 6 0 612.25 600 2 428.57 375 12.49

2 35 6 -1 518.37 600 -15.7 363.26 300 17.4
3 35 6 -2 424.49 375 11.65 306.12 300 1.9
4 35 6 -3 265.31 225 15.19 163.26 150 8.12
5 35 6 -4 285.71 300 -5 204.08 225 -10.25

2-4- Design of simulation experiments

After validating the model and proving its performance in predicting the laser cutting process,
15 simulations were done based on the response surface methodology (RSM). In these 15

of
simulations, three parameters, including the laser power, laser focal plane position, and laser
cutting speed, are considered as inputs data. The reason for selecting these three inputs is that
in experimental conditions, the laser source is capable of changing these three parameters.

ro
Also, top kerf width, bottom kerf width, alpha angle, and heat-affected zone are selected as
output parameters. Table 6 shows the variation ranges of input parameters. Also, the matrix
layout of the experiments, along with the values of the output results, is presented in Table 7.
-p
The star sign in Table 7 indicates that the cutting process is not performed under the relevant
conditions. The position of the focal length is demonstrated in Figure 6. Focal plane position
re
(FPP) was on the top of the workpiece where it was set as zero position. Also, the bottom and
top of FPP were considered negative and positive, respectively.
lP
na
ur

Figure 6- Diagram of the laser focal position.

Table 6- The variations range of input parameters in the designed experiments

Jo

Parameters Sign Unit -2 -1 0 1 2

Speed S [mm/s] 4 8 12 16 20
Power P [w] 30 35 40 45 50
Focal plane position FPP [mm] -3 -2 -1 0 1

12
Table 7- Matrix layout of experiments
Inputs Outputs

No.
Top Bottom Alpha
Speed Power FPP HAZ
Kerf kerf Angle
)mm/s( )W( )mm( )μm(
)μm( )μm( )°(
1 8 45 0 375 262.5 1.428571 2.013475
2 20 40 -1 300 225 1.333333 1.342624
3 12 40 -1 225 150 1.5 1.342624
4 12 40 1 * * * *
5 12 40 -3 187.5 187.5 1 0
6 16 35 -2 337.5 300 1.125 0.671404
7 16 35 0 300 150 2 2.683775
8 8 35 -2 525 375 1.4 2.683775
9 8 35 0 412.5 300 1.375 2.013475
10 16 45 0 337.5 225 1.5 2.013475
11 12 50 -1 262.5 150 1.75 2.013475
12 16 45 -2 300 375 0.8 -1.34262
13 4 40 -1 750 450 1.666667 5.355825
14 12 30 -1 150 150 1 0

of
15 8 45 -2 600 375 1.6 4.02199

3- Results and discussion

ro
In this section, statistical analysis and analysis of three inputs parameters including laser
power, laser focal plane position, and laser cutting speed on output parameters (top kerf
-p
width, bottom kerf width, the ratio of top kerf width to bottom kerf width, and heat-affected
zone) are presented. Each section also provides tables for the analysis of variance, regression
equation, RSM graphs, and standard curves for each output.
re
3-1- Top kerf width
Table 8 shows a modified variance analysis of the model for top kerf width. This table is
lP

obtained after removing ineffective parameters. P-value indicates the effect of each parameter
on the top kerf width. The calculation of the probability value is based on the assumption of
zero. For each parameter, if the P-value is less than 0.05, it means that the parameter is
effective on the top kerf width. It also indicates the significance of the model. Equation 6 is
na

the regression equation for the top kerf width. In this equation, and also the following
equation, some parameters are shown by their abbreviations such as speed (S), focal plane
position (FPP), and power (P). The values of R-sq and R-sq (adj) are the amount of data
ur

coverage by the regression equation. Figure 7 illustrates the curve of residuals distribution for
the top kerf width. It can be seen that the residual is scattered around the oblique line, and
shows a normal distribution; therefore, the final extracted regression model is also found a
Jo

suitable model for predicting and investigating the effects of the parameter.

13
Figure 7- Residuals curve for the response of the top kerf width

Table 8- The modified variance analysis for the top kerf width

Sum of Mean F P

of
Source Df
Squares Square Value value
Model 38045848 5 7609170 8.8446 0.0041
A-Speed 20073231 1 20073231 23.33234 0.0013

ro
B-Power 381489.2 1 381489.2 0.443428 0.5242
C-FPP 451547.9 1 451547.9 0.524862 0.4894
AC 1957151 1 1957151 2.274916 0.1699
A2 15622610 1 15622610 18.15911 0.0028
Residual
Total
6882545
44928392
8
13
R-sq: 84.6811% R-sq(adj): 75.1068%
-p
860318.1
re
lP
na
ur
Jo

14
(Top kerf) 1.35= 10111.18637-1441.953815 S + 30.88240787 P -1682.225816 FPP +
(6)
123.653719 S×FPP + 53.56615792 S2
The cutting speed and squared laser power have the highest effect on top kerf width,
respectively, and the focal length and laser power do not have much effect. Figure 8,
illustrates the effect of different parameters on the top kerf width. In this curve, the high slope
(positive or negative) indicates that the above parameter has a significant effect on the main
parameter (here, the top kerf width). Obviously, by increasing the laser power, the top kerf
increases, and FPP has an inverse effect. By increasing FPP, the top kerf reduces because the
area influenced via laser beam is further away than the close distance of FPP and workpiece
position.

of
ro
-p
re
Figure 8- The effect of different parameters on the top kerf width
lP
na
ur

)a( )b(
Figure 9- Top kerf RSM curve a) for cutting speed and laser focal plane position, b) for laser
power and laser speed
Jo

As shown in Figure 9a, the highest top kerf width occurs at a speed of 4 mm/s, and a focal
plane distance of -3 from the workpiece surface. In Figure 9a, the laser power is the average
range of 40 W. As the laser cutting speed increases from 14 mm/s, the top kerf width
decreases due to the interaction between the effects of cutting speed and laser power. It is
also found from Figure 9b that due to the interaction between the effect of cutting speed and
laser power, the highest top kerf width occurs at a speed of 4 mm/s and power of 50 W. This
phenomenon can be justified by the amount of the heat input. Equation 7 represents the
amount of heat input based on the laser power and speed [40].

15
 Heat input = P o w e r )7(
S peed

As the laser power increases, and the cutting speed decreases, the heat input increases, and
more of the workpiece melt, resulting in a higher top kerf width. In Figure 9b, the missing
parameter is the focal plane distances with an average range of -1. As the laser cutting speed
increases from 14 mm/s, the top kerf width has an increasing inverse trend that is justified by
the interaction between the cutting speed effect and the laser FPP effect.

3-2- The bottom kerf width

According to Table 9, the parameters of cutting speed and laser FPP have the greatest effect
on the bottom kerf width, respectively. Also, Equation 8 shows the regression relation for the
bottom kerf width.

of
Sum of Mean p-value
Source Df F-Value
Squares Square Prob > F
Model 6.425553 3 2.141851 1.287573 0.0315

ro
A-Speed 4.697476 1 4.697476 2.823885 0.0238
B-Power 0.154891 1 0.154891 0.093113 0.6665
C-FPP 1.573186 1 1.573186 0.94572 0.0437
Residual 16.6348 10 1.66348
Total 23.06035 13
R-sq: 73.3426% R-sq(adj): 60.9646%
-p
Table 9- The modified variance analysis for the bottom kerf width
re
(Bottom kerf) 0.39 = 9.050238377 – 0.135460379 S + 0.019678123 P – 0.366464717 FPP (8)
lP

Figure 10 illustrates the effect of different parameters on the bottom kerf width. The laser
FPP, cutting speed, and laser power are linearly correlated with the bottom kerf width, and
na

the maximum value for the bottom kerf width is 350 µm, which occurs at a cutting speed of 4
mm/s and a power of 20 W. According to Figure 10, the laser power has a direct relationship
with the bottom kerf width. As the laser power increases, the bottom kerf width increases by
17%. As can be seen from figure 10, the laser cutting speed is an effective parameter for the
ur

bottom kerf width. With reducing the laser cutting speed, the bottom kerf width will have an
increase due to the increased interaction time of the laser beam with the workpiece. It is
Jo

observed that by reducing the laser cutting speed from 20 mm/s to 4 mm/s, the bottom kerf
width increases around 45%.

16
Figure 10- The effect of different parameters on the bottom kerf width

As seen in Figure 10, it can be found that the focal plane position has a prominent
influence on the bottom kerf width. The farther the laser FPP from the workpiece surface
towards the depth of workpiece, the higher the bottom kerf width becomes. That is because

of
the focus of laser energy is positioned at the laser FPP, and the closer it is to the depth of the
workpiece, the bottom kerf width increases. Also, by increasing the laser FPP, the bottom

ro
kerf width increases by 30%.

3-3- Heat-affected zone

-p
Table 10 presents the modified variance analysis for the heat-affected zone. Laser cutting
speed and laser power have the most effect on the heat-affected zone, and the laser FPP can
almost be considered as an effective parameter.
re
Table 10- The modified variance analysis for heat-affected zone.
lP

Sum of Mean F P
Source Df
Squares Square Value value
Model 0.001917 4 0.000479 4.691289 0.0254
A-Speed 0.000773 1 0.000773 7.569327 0.0224
B-Power 0.000626 1 0.000626 6.124729 0.0353
na

C-FPP 0.000439 1 0.000439 4.292717 0.0682

AC 7.95E-05 1 7.95E-05 0.778385 0.4006
Residual 0.000919 9 0.000102
Cor Total 0.002836 13
R-sq: 67.5853% R-sq(adj): 53.1787%
ur

Also, the regression equation for the heat-affected zone is given in equation 9:
(HAZ) -0.22 = 0.29168069 + 0.002526117 S – 0.001250666 P -0.003339668FPP + 0.00078817 S
Jo

× FPP
(9)

Figure 11 demonstrates the effect of different input parameters on the heat-affected zone. The
results reveal that the cutting laser speed and laser power have a significant effect on the heat-
affected zone. By increasing FPP in the laser cutting process, the amount of HAZ area
reduces because the laser beam could make a wider zone treatment, and the concentration of
the laser beam on a particular zone is decreased.

17
 Figure 11- The effect of different parameters on the heat-affected zone

As shown in Figure 11, with increasing the laser power, the area of the heat-affected zone
increases by 38%, due to a rise in the energy entering into the surface of the workpiece.
According to Figure 11, it is seen that as the cutting speed decreases, the area of the heat-

of
affected zone increases by 40% due to the more absorbed energy by the workpiece at low
speeds. Conversely, at high cutting speeds, due to the less energy absorption, the heat-
affected zone decreases. Figure 11 displays that as the laser FPP moves deeper into the

ro
workpiece, the area of the heat-affected zone increases. By increasing the laser focal plane
position, the maximum amount of laser energy is approaching the bottom of the workpiece,
creating a more heat-affected zone. Using equation 11, the RSM curves are plotted for the
heat-affected zone in Figure 12.
-p
re
lP
na

Figure 12- The HAZ response surface diagram of the focal plane position and cutting speed
ur

It can be seen from Figure 12 that both the laser FPP and the cutting speed parameters are
effective in the area of heat-affected zone and the lowest heat-affected zone also occurs at the
highest FPP (+1) and the maximum laser cutting speed (20 mm/s). As shown, by increasing
Jo

the cutting speed at the focal plane position of -3, the heat-affected zone does not change
significantly, but at the FPP of +1, with increasing cutting speed, the heat-affected zone
decreases. That is because at the FPP of -3, all laser energy concentration is within the
workpiece, and the cutting speed does not have much effect on the heat-affected zone, in
contrast at the focal plane position of +1, the main laser energy concentration is 1 mm above
the workpiece surface, and therefore, increasing the laser cutting speed reduces the heat
interaction with the surface of the workpiece, and consequently decreases the heat-affected
zone.

18
 3-4- Kerf taper angle

Table 11 shows the modified variance analysis for the alpha angle. According to P-value,
cutting speed is the most effective parameter among the main parameters. And the squared
cutting speed (in equation 10) is the single parameter that affects the value of the alpha angle.
Table 11- The modified variance analysis for the alpha angle.
Sum of Mean F P
Source Df
Squares Square Value Value
Model 12.52615 7 1.789451 4.2824 0.0481
A-Speed 4.288691 1 4.288691 10.26342 0.0185
B-Power 1.273761 1 1.273761 3.048283 0.1314
C-FPP 0.560442 1 0.560442 1.341213 0.2908
AB 0.08072 1 0.08072 0.193174 0.6757
AC 1.857274 1 1.857274 4.444711 0.0796
BC 0.454821 1 0.454821 1.088448 0.3370
A^2 3.514305 1 3.514305 8.410212 0.0273

of
Residual 2.50717 6 0.417862
Cor Total 15.03332 13
R-sq: 83.3226% R-sq(adj): 63.8656%

ro
(Alpha angle) 0.82 = 2.452636097 – 0.417816779 S + 0.069012384 P + 0.68302616 FPP –
0.005022454 S ×P + 0.120457279 S ×FPP – 0.047687565 P × FPP + 0.025405833 S2
(10)
-p
Figure 13 illustrates the effect of different input parameters on the alpha angle. As can be
seen, the laser cutting speed has a weighty effect on the alpha angle.
re
lP
na

Figure 13- The effect of different parameters on the alpha angle

ur

According to the results presented in Table 11 and Figure 13, the laser power and laser focal
plane position have a direct and linear relationship with the alpha angle, and the laser cutting
Jo

speed has a non-linear relationship with the alpha angle. Also, according to Table 11, the
laser power does not affect the alpha angle value. However, according to Figure 13, as the
laser power increases, the alpha angle increases too. The reason is that an increase in the laser
energy, increases the difference between the top and bottom kerf widths. Figure 13 reveals
that the lower laser cutting speed, the more alpha angle increases due to the increase of the
laser interaction time with the workpiece. The laser focal plane position does not affect the
alpha angle, but Figure 13 shows that as the laser focal length goes to the depth of the
workpiece, the alpha angle increases. According to Table 11 and equation 12, which is the
regression relation for the alpha angle, the response surface is plotted in Figure 14.

19
 Figure 14- The Taper angle diagram of a response surface for cutting speed and focal plane position

of
Figure 14 represents that the highest value of the alpha angle occurs at the cutting speed of 4
mm/s, and laser FPP of -3 mm. The lowest amount of the alpha angle occurs at cutting speed
of 20 mm/s and laser FPP of 3 mm. At the focal plane position of +1, with increasing the

ro
cutting speed up to 14 mm/s, the alpha angle decreases, and at cutting speed of 14 mm/s or
higher, the alpha angle increases due to the interaction of the laser effects between the cutting
speed and the laser power (average range of 40 W).

4- Optimization
-p
In this study, after the simulations, according to the model given by the design of
re
experiments, the optimization of the process is done by a design expert software, in which the
outputs of the simulation process are optimized. The optimization aims are the highest top
kerf width, the lowest bottom kerf width, the lowest alpha angle, the kerf ratio of 1, and the
lP

lowest heat-affected zone, respectively. In this process, using the optimization results, the
simulation of the polycarbonate laser cutting process is repeated. Five optimization criteria
are modified by varying the importance of each of the optimal parameters to create a suitable
model for simulating the sample, as shown in Table 12. After optimizing the experiments, the
na

simulation results, and the optimization results are compared with the design expert program,
and the error percentage for each output response is given in Table 13. According to the
obtained results, the error percentage between the simulated samples, and the samples
extracted from the software is very low, and this shows the excellent performance of the
ur

simulation. Also, Figure 15 shows the over plot, which is the integration of contour plots of
each response in the top view of each other. In each part, areas of less importance are shown
in gray. The remaining yellow area also represents the final desired region of the input
Jo

parameters. For example, in Figure 15a, the center of the diagram shows proper output
parameters influenced by speed and FPP. As mentioned in the previous sections, the
interaction of parameters is an important part of this research, and by these figures, the
concept of optimized parameters is shown better.
Table 12- Constraints and criteria of input parameters and responses
Parameter/Response Goal Lower Upper Importance
Scanning speed in range 4 20 ---
Parameters Laser power in range 30 50 ---
Focal plane position in range -3 1 ---
R

n
o
e

e
s

Criteria 1 Top Kerf Maximize 150 750 1

20
 Bottom kerf Minimize 150 450 1
Ratio 1 0.8 2 4
HAZ Minimize 262.5 712.5 5
Alpha Angle Minimize 0 5.35 1

Criteria 2 Top Kerf Maximize 150 750 1

Bottom kerf Minimize 150 450 3
Ratio 1 0.8 2 5
HAZ Minimize 262.5 712.5 3
Alpha Angle Minimize 0 5.35 3

Criteria 3 Top Kerf Maximize 150 750 3

Bottom kerf Minimize 150 450 4
HAZ Minimize 262.5 712.5 4
Alpha Angle Minimize 0 5.35 2

Criteria 4 Top Kerf Maximize 150 750 4

Bottom kerf Minimize 150 450 4
HAZ Minimize 262.5 712.5 4
Alpha Angle Minimize 0 5.35 3
Criteria 4 Top Kerf Maximize 150 750 5
Bottom kerf Minimize 150 450 5

of
HAZ Minimize 262.5 712.5 5
Alpha Angle Minimize 0 5.35 5

ro
Table 13- Optimum prediction results and experimental validation
Solution

Input parameters Output parameters

Composite

1
Speed
)mm/s(

6.84
Power
)W(

30
FPP
)mm(

1
desirability

0.9097
Abaqus
Predicted
375
-p
Top Kerf
)μm(

361.858
Bottom kerf

225
)μm(

230.791
Ratio

1.6666
0.99
279
HAZ
)μm(

289.14
2.68
2.37
Alpha Angle
)°(
re
Error% 3.5 2.5 43.7 3.6 11.6
Abaqus 300 150 2 324 2.68
2 20 33.03 -1.85 0.9192 Predicted 304.781 185.61 1.7 342.348 2.57
Error% 2 23.3 15 5.6 4.1
Abaqus 225 150 1.5 285 1.34
lP

3 12.87 30 1 0.9167 Predicted 198.363 177.073 1.12 273.421 0.97

Error% 11.8 18 25.3 39.1 27.6
Abaqus 175 175 1 268 0.68
4 12.40 30 1 0.9108 Predicted 196.461 180.938 1.08 274.662 0.91
Error% 12.2 3.4 8 2.5 33.8
Abaqus 175 175 1 265 0.72
na

5 12.08 30 1 0.9065 Predicted 196.728 183.648 1.07 275.508 0.80

Error% 14.4 4.9 7 3.4 11.1
ur
Jo

(a)

21
 (b)
Figure 15 Overlay plot a) Laser FFP and cutting speed b) Laser power and cutting speed

of
5- Conclusion

ro
In this study, a 3.2 mm thick polycarbonate sheet was selected. The laser used was a
continuous mode low power CO2 laser. The input parameters of the simulation included laser
power, cutting speed, and laser focal plane position. Gas pressure (3 Bar) was also assumed
-p
for the polycarbonate material. The parameters of top kerf width, and the bottom kerf width,
heat-affected zone, the ratio of the top kerf width to the bottom kerf width and alpha angle
were considered as simulation output responses. The difference between the results of
re
numerical simulation with experiments under similar conditions was less than 12%, which
indicates appropriate modeling accuracy.
lP

The results showed that the highest top kerf width occurs at the cutting speed of 4 mm/s and
the focal plane distance of -3 from the workpiece, and the lowest kerf width is obtained when
the focal plane position is at the deepest part of the workpiece at the position of 3 mm. It was
shown that the laser cutting speed has a significant effect on the ratio of top kerf width to
na

bottom kerf width, and by increasing the laser cutting speed from 4 to 20 mm/s, this ratio
increases from 0 to 875.1 µm. Furthermore, by increasing the laser cutting speed from 4 to 20
mm/s, and decreasing the laser power from 50 to 20 W, the heat-affected zone decreases.
ur

Declaration of interests
Jo

The authors declare that they have no known competing financial interests or personal
relationships that could have appeared to influence the work reported in this paper.

22
References
[1] S. M. Karazi, M. Moradi, and K. Y. Benyounis, Statistical and Numerical Approaches for Modeling and
Optimizing Laser Micromachining Process-Review. Elsevier Ltd., 2019.
[2] M. Moradi, O. Mehrabi, T. Azdast, and K. Y. Benyounis, “Enhancement of low power CO2 laser cutting
process for injection molded polycarbonate,” Opt. Laser Technol., vol. 96, pp. 208–218, 2017, DOI:
10.1016/j.optlastec.2017.05.022.
[3] B. El Aoud, M. Boujelbene, E. Bayraktar, S. Ben Salem, and I. Miskioglu, “Studying Effect of CO 2
Laser Cutting Parameters of Titanium Alloy on Heat-Affected Zone and Kerf Width Using the Taguchi
Method,” vol. 6, pp. 143–150, 2018, DOI: 10.1007/978-3-319-63408-1.
[4] M. Kurt, Y. Kaynak, and E. Bagci, “Dimensional analyses and surface quality of the laser cutting
process for engineering plastics,” pp. 259–267, 2009, DOI: 10.1007/s00170-008-1468-7.
[5] S. Marimuthu, J. Dunleavey, Y. Liu, M. Antar, and B. Smith, “Laser cutting of aluminium-alumina
metal matrix composite,” Opt. Laser Technol., vol. 117, no. February, pp. 251–259, 2019, DOI:

of
10.1016/j.optlastec.2019.04.029.
[6] M. Sharifi and M. Akbari, “Experimental investigation of the effect of process parameters on cutting

ro
region temperature and cutting edge quality in laser cutting of AL6061T6 alloy,” Optik (Stuttg)., vol.
184, pp. 457–463, 2019, DOI: 10.1016/j.ijleo.2019.04.105.
[7] M. S. Shaikh Mohammad Meiabadi, A. Kazerooni, M. Moradi, and M. J. Torkamany, “Laser-assisted
joining of St12 to polycarbonate: Experimental study and numerical simulation,” Optik (Stuttg)., p.

[8]
164151, 2019, DOI: 10.1016/j.ijleo.2019.164151. -p
M. H. El-Hofy, and H. El-Hofy, “Laser beam machining of carbon fiber-reinforced composites: a
review,” Int. J. Adv. Manuf. Technol., vol. 101, no. 9–12, pp. 2965–2975, 2019, DOI: 10.1007/s00170-
re
018-2978-6.
[9] A. M. Varsi, and A. H. Shaikh, “ Experimental and statistical study on kerf taper angle during CO 2
laser cutting of thermoplastic material ,” J. Laser Appl., vol. 31, no. 3, p. 032010, 2019, DOI:
lP

10.2351/1.5087846.
[10] E. Haddadi, M. Moradi, A. Karimzad Ghavidel, A. Karimzad Ghavidel, and S. Meiabadi, “Experimental
and parametric evaluation of cut quality characteristics in CO2 laser cutting of polystyrene,” Optik
(Stuttg)., vol. 184, no. February, pp. 103–114, 2019, DOI: 10.1016/j.ijleo.2019.03.040.
na

[11] X. Chen, T. Li, K. zhai, Z. Hu, and M. Zhou, “Using orthogonal experimental method optimizing the
surface quality of CO2 laser cutting process for PMMA microchannels,” Int. J. Adv. Manuf. Technol.,
vol. 88, no. 9–12, pp. 2727–2733, 2017, DOI: 10.1007/s00170-016-8887-7.
[12] S. J. S. Al-Qaisy, and C. Karataş, “Laser cutting of high-density poly ethylene (Hdpe) pipes Pe100 using
ur

Co2 gas laser,” Int. J. Eng. Adv. Technol., vol. 9, no. 1, pp. 6193–6199, 2019, DOI:
10.35940/ijeat.A1783.109119.
[13] C. M. Gruescu, C. L. Ionescu, I. Nicoara, and A. Lovasz, “Experimental optimization of process
parameters in laser cutting of polycarbonate gears,” Mechanika, vol. 18, no. 2, pp. 233–238, 2012, DOI:
Jo

10.5755/j01.mech.18.2.1561.
[14] A. M. Varsi, and A. H. Shaikh, “ Developing an algorithm for predicting depth as well as a number of
passes during CO 2 laser machining on thermoplastic material ,” J. Laser Appl., vol. 30, no. 4, p.
042007, 2018, DOI: 10.2351/1.5048055.
[15] S. Sihn, J. Pitz, R. H. Gerzeski, A. K. Roy, and J. P. Vernon, “Experimentally-validated computational
model for temperature evolution within laser heated fiber-reinforced polymer matrix composites,”
Compos. Struct., vol. 207, pp. 966–973, 2019, DOI: 10.1016/j.compstruct.2018.09.041.
[16] T. Ohkubo, M. Tsukamoto, and Y. Sato, “Numerical simulation of laser beam cutting of carbon fiber
reinforced plastics,” Phys. Procedia, vol. 56, no. C, pp. 1165–1170, 2014, DOI:
10.1016/j.phpro.2014.08.031.

23
[17] B. S. Yilbas, S. S. Akhtar, C. Karatas, H. Ali, and N. Abu-Dheir, “Laser Cutting of Holes in Inconel 803
Alloy and Analysis of Thermal Stress Field,” Mach. Sci. Technol., vol. 23, no. 1, pp. 95–117, 2019,
DOI: 10.1080/10910344.2018.1486412.
[18] E. H. Amara et al., “Simulation by temperature gradient adaption of wavelength effect in metal laser
cutting,” J. Laser Appl., vol. 29, no. 2, p. 022209, 2017, DOI: 10.2351/1.4983264.
[19] V. Pata, L. Sýkorová, O. Šuba, and M. KubiŠová, “Simulation of the Transient Temperature Field when
Laser Machining Polymeric Materials,” Key Eng. Mater., vol. 686, no. 1, pp. 246–251, 2016, DOI:
10.4028/www.scientific.net/KEM.686.246.
[20] Y. Ayed, G. Germain, W. Ben Salem, and H. Hamdi, “Experimental and numerical study of laser-
assisted machining of Ti6Al4V titanium alloy,” Finite Elem. Anal. Des., vol. 92, pp. 72–79, 2014, DOI:
10.1016/j.finel.2014.08.006.
[21] J. Meško, A. Zrak, R. Nigrovič, and M. Ţmindák, “Finite element simulation of laser cutting process of
steel sheet,” MATEC Web Conf., vol. 157, pp. 2–7, 2018, DOI: 10.1051/matecconf/201815702030.
[22] K. Kheloufi, E. H. Amara, and A. Benzaoui, “Numerical simulation of transient three-dimensional
temperature and kerf formation in laser fusion cutting,” J. Heat Transfer, vol. 137, no. 11, pp. 1–9,

of
2015, DOI: 10.1115/1.4030658.
[23] A. Ottoa and M. Schmidta, “Towards a universal numerical simulation model for laser material
processing,” Phys. Procedia, vol. 5, no. PART 1, pp. 35–46, 2010, DOI: 10.1016/j.phpro.2010.08.120.

ro
[24] M. Benton, M. R. Hossan, P. R. Konari, and S. Gamagedara, “Effect of process parameters and material
properties on laser micromachining of microchannels,” Micromachines, vol. 10, no. 2, 2019, DOI:
10.3390/mi10020123.
[25]
-p
G. Le, A. Hoang, D. Nguyen, M. Nguyen, and Q. Nguyen, “A Method to Determine Machining
Parameters for Laser Cutting Machine Using Numerical Simulation,” vol. 2, no. 9, pp. 65–68, 2018.
[26] M. Moradi, and E. Golchin, “Investigation on the effects of process parameters on laser percussion
re
drilling using finite element methodology; statistical modeling and optimization,” Lat. Am. J. Solids
Struct., vol. 14, no. 3, pp. 464–484, 2017, DOI: 10.1590/1679-78253247.
[27] V. Tangwarodomnukun, and S. Mekloy, “Temperature field modeling and cut formation in laser
lP

micromachining of silicon in ice layer,” J. Mater. Process. Technol., vol. 271, no. October 2018, pp.
202–213, 2019, DOI: 10.1016/j.jmatprotec.2019.04.006.
[28] S. Choi, and K. Y. Jhang, “Numerical study on thermal stress cutting of silicon wafers using two-line
laser beams,” J. Mech. Sci. Technol., vol. 33, no. 8, pp. 3621–3627, 2019, DOI: 10.1007/s12206-019-
na

0702-6.
[29] Paul, B.K., Ahmed, K., Vigneswaran, D., Ahmed, F., Roy, S., and Abbott, D. Quasi-photonic crystal
fiber-based spectroscopic chemical sensor in the terahertz spectrum: Design and analysis. IEEE Sensors
Journal, 18(24), pp.9948-9954, 2018, DOI: 10.1109/JSEN.2018.2872892.
ur

[30] Jabin, M.A., Ahmed, K., Rana, M.J., Paul, B.K., Islam, M., Vigneswaran, D., and Uddin, M.S. Surface
plasmon resonance-based titanium coated biosensor for cancer cell detection. IEEE Photonics Journal,
11(4), pp.1-10, 2019, DOI: 10.1109/JPHOT.2019.2924825.
Jo

[31] Ahmed, K., Haque, M.J., Jabin, M.A., Paul, B.K., Amiri, I.S., and Yupapin, P., Tetra-core surface
plasmon resonance based biosensor for alcohol sensing. Physica B: Condensed Matter, 570, pp.48-52,
2019, DOI: 10.1016/j.physb.2019.05.047.
[32] Ahmed, K., Paul, B.K., Jabin, M.A., and Biswas, B., FEM analysis of birefringence, dispersion, and
nonlinearity of graphene-coated photonic crystal fiber. Ceramics International, 45(12), pp.15343-15347,
2019, DOI: 10.1016/j.ceramint.2019.05.027.
[33] Paul, B.K., Chakma, S., Khalek, M.A., and Ahmed, K., Silicon nanocrystal filled ellipse core-based
quasi photonic crystal fiber with birefringence and very high nonlinearity. Chinese Journal of Physics,
56(6), pp.2782-2788, 2018, DOI: 10.1016/j.cjph.2018.09.030.
[34] Thenmozhi, H., Rajan, M.M., and Ahmed, K., D-shaped PCF sensor based on SPR for the detection of

24
 carcinogenic agents in food and cosmetics. optik, 180, pp.264-270, 2019, DOI:
10.1016/j.ijleo.2018.11.098.
[35] Li, M., Gan, G., Zhang, Y. and Yang, X, “Thermal defect characterization and strain distribution of
CFRP laminate with open hole following fiber laser cutting process,” Optics & Laser Technology, 2020,
p.105891, DOI: 10.1016/j.optlastec.2019.105891, 122.
[36] Kaselouris, E., Baroutsos, A., Papadoulis, T., Papadogiannis, N.A., Tatarakis, M. and Dimitriou, V.,
2020. A Study on the Influence of Laser Parameters on Laser-Assisted Machining of Aisi H-13 Steel. In
Key Engineering Materials, Vol. 827, pp. 92-97, DOI: 10.4028/www.scientific.net/KEM.827.92.
[37] R. Akarapu, B. Q. Li, and A. Segall, “A Thermal Stress and Failure Model for Laser Cutting and
Forming Operations,” vol. 4, no. October, pp. 51–62, 2004, DOI: 10.1361/15477020420756.
[38] M. J. Kim, “3D finite element analysis of evaporative laser cutting,” vol. 29, pp. 938–954, 2005, DOI:
10.1016/j.apm.2005.02.015.
[39] M. Moradi, M. Karami Moghadam, M. Shamsborhan, M. Bodaghi, H. Falavandi, Post Processing of
FDM 3D Printed Poly Lactic Acid Parts by Laser Beam Cutting, polymers, 2020.

of
[40] M. Moradi, and M. Karami moghadam, “High power diode laser surface hardening of AISI 4130 ;
statistical modeling and optimization,” Opt. Laser Technol., vol. 111, no. October 2018, pp. 554–570,
2019, DOI: 10.1016/j.optlastec.2018.10.043.

ro
-p
re
lP
na
ur
Jo

25